Autosense; Unit Attention Condition - Seagate SCSI Interface Product Manual

SCSI Interface Product Manual, Rev. J
status on a subsequent command, but not both. Notification of an error condition encountered after command
completion shall be returned only to the initiator that sent the affected task or tasks.
Asynchronous event reports may be used to notify devices that a system resource has become available. If a
logical unit uses this method of reporting, the sense key in the AER sense data shall be set to Unit Attention.
4.6.4.2

Autosense

Autosense is the automatic return of sense data to the application client coincident with the completion of an
SCSI-3 command under the conditions described below. The return of sense data in this way is equivalent to
an explicit command from the application client requesting sense data immediately after being notified that an
ACA condition has occurred. Inclusion of autosense support in an SCSI-3 protocol standard is optional.
As specified in Section 4, the application client may request autosense service for any SCSI command. If sup-
ported by the protocol and logical unit and requested by the application client, the device server shall only
return sense data in this manner coincident with the completion of a command with a status of Check Condi-
tion. After autosense data is sent, the sense data and the CA (NACA=0), if any, shall then be cleared.
Autosense shall not affect ACA (NACA=1), see Section 4.6.1.
Protocol standards that support autosense shall require an autosense implementation to:
a. Notify the logical unit when autosense data has been requested for a command; and
b. Inform the application client when autosense data has been returned upon command completion (see Sec-
tion 4.0).
It is not an error for the application client to request the automatic return of sense data when autosense is not
supported by the SCSI-3 protocol or logical unit implementation. If the application client requested the return of
sense data through the autosense facility and the protocol service layer does not support this feature, then the
confirmation returned by the initiator's service delivery port should indicate that no sense data was returned. If
the protocol service layer supports autosense but the logical unit does not, then the target should indicate that
no sense data was returned. In either case, sense information shall be preserved and the application client
may issue a command to retrieve it.
4.6.5

Unit attention condition

The drive sets up the Unit Attention condition when it stores (within itself) a Unit Attention condition flag for
each device on the SCSI bus having an initiator relationship with the drive, and this Unit Attention condition
persists for each initiator until the condition is cleared (flag negated) by each initiator individually. The Unit
Attention condition results when one of the following events occur:
1. A power-on sequence occurs.
2. A reset is generated internally by the drive (caused by a power glitch).
3. A Bus Device Reset message causes the drive to reset itself.
4. The RESET I/O line resets the drive.
5. An initiator changes one or more of the Mode Select parameters in the drive (these changes could affect
one or more of the other initiators).
6. The inquiry data has been changed.
7. The mode parameters in effect for an initiator have been restored from non-volatile memory.
8. An event occurs that requires the attention of the initiator.
9. A Clear Queue message is received.
10. The Log parameters are changed. Unit Attention Condition is posted for all initiators in the system other
than the one that changed the Log Parameters.
11. Tasks for this initiator were changed by another initiator.
12. The version or level of microcode has changed.
The Unit Attention Parameters page (page 00h, bit 4 of byte 2) of the Mode Select command controls whether
or not a Check Condition Status is to be reported to affected initiators when a Unit Attention condition exists.
See Table 101.
69